AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

winsys/amdgpu: merge 2 loops iterating over sparse BOs in amdgpu_cs_submit_ib

Browse Source 
The first loop is inside amdgpu_add_sparse_backing_buffers. The second loop
iterates over sparse BOs to update sequence numbers.

Each loop has to reload the whole BO list from the L2 cache or higher, so we
do it twice. By merging the loops, we only load the BO list from the L2
cache once.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27408>
This commit is contained in:
Marek Olšák
2024-01-30 10:52:42 -05:00
parent 32625c7903
commit d704f92b13
1 changed files with 48 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/gallium/winsys/amdgpu/drm/amdgpu_cs.c
+48 -64
View File
@@ -1191,23 +1191,6 @@ static void amdgpu_set_bo_seq_no(unsigned queue_index, struct amdgpu_winsys_bo *
bo->fences.valid_fence_mask |= BITFIELD_BIT(queue_index);
}
static void amdgpu_add_bo_fences_to_dependencies(struct amdgpu_winsys *ws, unsigned queue_index,
struct amdgpu_seq_no_fences *dependencies,
uint_seq_no new_queue_seq_no,
struct amdgpu_buffer_list *list)
{
struct amdgpu_cs_buffer *buffers = list->buffers;
unsigned num_buffers = list->num_buffers;
for (unsigned i = 0; i < num_buffers; i++) {
struct amdgpu_cs_buffer *buffer = &buffers[i];
struct amdgpu_winsys_bo *bo = buffer->bo;
amdgpu_add_fences_to_dependencies(ws, queue_index, dependencies, bo, buffer->usage);
amdgpu_set_bo_seq_no(queue_index, bo, new_queue_seq_no);
}
}
static void amdgpu_cs_add_syncobj_signal(struct radeon_cmdbuf *rws,
struct pipe_fence_handle *fence)
{
@@ -1217,43 +1200,6 @@ static void amdgpu_cs_add_syncobj_signal(struct radeon_cmdbuf *rws,
add_fence_to_list(&cs->syncobj_to_signal, (struct amdgpu_fence*)fence);
}
/* Add backing of sparse buffers to the buffer list.
*
* This is done late, during submission, to keep the buffer list short before
* submit, and to avoid managing fences for the backing buffers.
*/
static bool amdgpu_add_sparse_backing_buffers(struct amdgpu_cs_context *cs)
{
struct amdgpu_cs_buffer *buffers = cs->buffer_lists[AMDGPU_BO_SPARSE].buffers;
unsigned num_sparse_buffers = cs->buffer_lists[AMDGPU_BO_SPARSE].num_buffers;
for (unsigned i = 0; i < num_sparse_buffers; ++i) {
struct amdgpu_cs_buffer *buffer = &buffers[i];
struct amdgpu_bo_sparse *bo = get_sparse_bo(buffer->bo);
simple_mtx_lock(&bo->commit_lock);
list_for_each_entry(struct amdgpu_sparse_backing, backing, &bo->backing, list) {
/* We can directly add the buffer here, because we know that each
* backing buffer occurs only once.
*/
struct amdgpu_cs_buffer *real_buffer =
amdgpu_do_add_buffer(cs, &backing->bo->b, &cs->buffer_lists[AMDGPU_BO_REAL]);
if (!real_buffer) {
fprintf(stderr, "%s: failed to add buffer\n", __func__);
simple_mtx_unlock(&bo->commit_lock);
return false;
}
real_buffer->usage = buffer->usage;
}
simple_mtx_unlock(&bo->commit_lock);
}
return true;
}
static void amdgpu_cs_submit_ib(void *job, void *gdata, int thread_index)
{
struct amdgpu_cs *acs = (struct amdgpu_cs*)job;
@@ -1348,10 +1294,54 @@ static void amdgpu_cs_submit_ib(void *job, void *gdata, int thread_index)
real_buffer->usage |= buffer->usage & ~RADEON_USAGE_SYNCHRONIZED;
}
amdgpu_add_bo_fences_to_dependencies(ws, queue_index, &seq_no_dependencies, next_seq_no,
&cs->buffer_lists[AMDGPU_BO_SPARSE]);
amdgpu_add_bo_fences_to_dependencies(ws, queue_index, &seq_no_dependencies, next_seq_no,
&cs->buffer_lists[AMDGPU_BO_REAL]);
/* Sparse BOs: Add fence dependencies, update seq_no in BOs, add real buffers. */
unsigned num_real_buffers_except_sparse = cs->buffer_lists[AMDGPU_BO_REAL].num_buffers;
struct amdgpu_cs_buffer *sparse_buffers = cs->buffer_lists[AMDGPU_BO_SPARSE].buffers;
unsigned num_sparse_buffers = cs->buffer_lists[AMDGPU_BO_SPARSE].num_buffers;
for (unsigned i = 0; i < num_sparse_buffers; i++) {
struct amdgpu_cs_buffer *buffer = &sparse_buffers[i];
struct amdgpu_winsys_bo *bo = buffer->bo;
amdgpu_add_fences_to_dependencies(ws, queue_index, &seq_no_dependencies, bo, buffer->usage);
amdgpu_set_bo_seq_no(queue_index, bo, next_seq_no);
/* Add backing buffers of sparse buffers to the buffer list.
*
* This is done late, during submission, to keep the buffer list short before
* submit, and to avoid managing fences for the backing buffers.
*/
struct amdgpu_bo_sparse *sparse_bo = get_sparse_bo(buffer->bo);
simple_mtx_lock(&sparse_bo->commit_lock);
list_for_each_entry(struct amdgpu_sparse_backing, backing, &sparse_bo->backing, list) {
/* We can directly add the buffer here, because we know that each
* backing buffer occurs only once.
*/
struct amdgpu_cs_buffer *real_buffer =
amdgpu_do_add_buffer(cs, &backing->bo->b, &cs->buffer_lists[AMDGPU_BO_REAL]);
if (!real_buffer) {
fprintf(stderr, "%s: failed to add sparse backing buffer\n", __func__);
simple_mtx_unlock(&sparse_bo->commit_lock);
r = -ENOMEM;
goto cleanup;
}
real_buffer->usage = buffer->usage;
}
simple_mtx_unlock(&sparse_bo->commit_lock);
}
/* Real BOs: Add fence dependencies, update seq_no in BOs except sparse backing BOs. */
struct amdgpu_cs_buffer *real_buffers = cs->buffer_lists[AMDGPU_BO_REAL].buffers;
for (unsigned i = 0; i < num_real_buffers_except_sparse; i++) {
struct amdgpu_cs_buffer *buffer = &real_buffers[i];
struct amdgpu_winsys_bo *bo = buffer->bo;
amdgpu_add_fences_to_dependencies(ws, queue_index, &seq_no_dependencies, bo, buffer->usage);
amdgpu_set_bo_seq_no(queue_index, bo, next_seq_no);
}
#if 0 /* Debug code. */
printf("submit queue=%u, seq_no=%u\n", acs->queue_index, next_seq_no);
@@ -1426,12 +1416,6 @@ static void amdgpu_cs_submit_ib(void *job, void *gdata, int thread_index)
} else
#endif
{
if (!amdgpu_add_sparse_backing_buffers(cs)) {
fprintf(stderr, "amdgpu: amdgpu_add_sparse_backing_buffers failed\n");
r = -ENOMEM;
goto cleanup;
}
struct amdgpu_cs_buffer *real_buffers = cs->buffer_lists[AMDGPU_BO_REAL].buffers;
unsigned num_real_buffers = cs->buffer_lists[AMDGPU_BO_REAL].num_buffers;
bo_list = alloca((num_real_buffers + 2) * sizeof(struct drm_amdgpu_bo_list_entry));